(602af) Looking for Optimal Conditions to Manufacture Activated Carbon From Barley Husk by Single- or Dual-Optimization

Activated carbon has a strong affinity for binding organic or inorganic pollutants from aqueous solutions due to its high specific surface area and the possibility of incorporating functional groups on surface. For these reasons, activated carbons are commonly used for wastewater treatment. Despite of the above-mentioned advantages, the use of commercial activated carbon as adsorbent is often limited due to economical reasons, and many researchers have paid attention in the use of alternative precursors, such as agro-waste materials for activated carbon production. In this research, the manufacturing of activated carbon from barley husk (BH) by chemical activation with zinc chloride was optimized using a factorial design with two responses (yield and iodine number) and three factors (activation temperature, activation time and impregnation ratio). Optimal conditions for manufacturing barley husk activated carbon (BHAC) were obtained using a 2³ factorial design followed by a central composite design by adding axial points to the initial design to explore the space curvature. Single or dual optimization was performed to determine the optimal conditions which improve surface area and yield.

Findings demonstrate that a linear model was not appropriate to represent the design space based on the ANOVA (at a confidence level of 99%) and, therefore, a central composite design was used to fit a quadratic model to the experimental data. According to the ANOVA for the quadratic model of iodine number, the most significant individual factors were temperature and impregnation ratio, double interaction between temperature and impregnation ratio, and the quadratic term of temperature (similar significant factors were determined for yield response except that the quadratic term was the impregnation ratio instead of temperature). It is important to mention that activation time was not significant and, for this reason, it was not considered in the quadratic model but for economical reasons time was set equal to 20 minutes.

Single optimization using the quadratic model of each response gives different optimal conditions depending on the response, for instance: maximal iodine number (1046.26±105.52 mg/g) was predicted at 530°C and impregnation ratio of 1.5 g ZnCl₂/g BH, while for maximal yield (61.58±3.32%) were found at 300°C and impregnation ratio of 0.76 g ZnCl₂/g BH. Taking into account that maximal responses were found at different conditions and both responses play an important role in manufacturing activated carbon from lignocelullosic precursors, both responses were optimized simultaneously by using the desirability functions approach included in Design expert software.

After simultaneous dual optimization, the maximal desirability function value (0.779) was obtained at 436°C, activation time of 20 minutes, and impregnation ratio of 1.1 g ZnCl₂/g BH. At these conditions, predicted responses for iodine number and yield were 829.58±78.30 mg/g and 46.82±2.64%, respectively, whereas experimental tests achieved 901.86 mg/g and 48.48% (values within the predicted interval). Obtained iodine number (901.86 mg/g) is similar or greater than those reported by other researchers dealing with lignocelullosic precursors. Correspondingly, BHAC produced at the optimal conditions (assessed by simultaneous dual optimization) reached a surface area (811.44 m²/g) similar to commercial activated carbons and lignocelullosic-based activated carbons.

In summary, simultaneous dual optimization seems to be better than single optimization due to the fact that both yield and surface area are very important properties when manufacturing activated carbon from lignocelullosic precursors.